REACTIONS IN MICROEMULSION MEDIA. BOROHYDRIDE REDUCTION OF MONO- AND DICARBONYL COMPOUNDS

Article abstract of DOI:10.1016/S0040-4020(01)96887-5

Microemulsions were prepared at 26 deg C from mixtures of hexanes (O), a 50:1 (w/w) solution (W) of 0.1 M KOH-NaBH4, and a 1.23:1 (w/w) mixture (S) of hexadecyltrimethylammonium bromide (HTABr) and 1-butanol.A pseudoternary phase map contained a significant microemulsion (μE) region, and μE's A and B (60:35:5 and 20:10:70 S:W:O, respectively) were used for reduction of several monocarbonyl compounds , an α,β-unsaturated ketone , and a diketone <4-(4'-benzoylphenyl)-2-butanone (7a)> at 26 deg C.For comparison purposes, reductions were also performed in aqueous 2-propanol (2-PrOH A and 2-PrOH B) prepared by the substitution of 2-propanol for the S and O components of μE's A and B.Generally, the reductions were slightly faster in the microemulsion media than in the corresponding aqueous 2-propanol media.The significantly slower reduction of 4a relative to that of 3a in μE B indicated that the interphase is the reactive site.With enone 6a, the influence of microemulsions on the competition between 1,2- and 1,4-reduction was determined.In μE's A and B there was 8percent and 11percent 1,4-reduction, respectively, whereas in 2-PrOH A and B there was only a trace.With diketone 7a, the reactivity of the aromatic carbonyl group relative to that of the aliphatic carbonyl group increased on going from 2-PrOH A and B to μE's A and B, respectively.For the sodium borohydride reduction of ketones, microemulsion catalysis is more effective than phase transfer catalysis or the use of a tetraalkylammonium borohydride in a hydrocarbon solvent.

Full text of DOI:10.1016/S0040-4020(01)96887-5

Tetrahedron Vol.
U.S.A.
No. 14,
2691-2698, 1984
$3.00
in
Press Ltd.
REACTIONS IN MICROEMULSION MEDIA. BOROHYDRIDE
REDUCTION OF MONO- AND DICARBONYL COMP OUNDS’
D
AVID A. JAEG ER,* MARY D ARLENE W ARD and CRAIG A. MARTIN
Department of Chemistry, University of Wyoming, Laramie, WY 82071, U.S.A.
(Received in USA 16 September 1983)
were prepared at 26” from mixtures of hexanes (0), a
1 (w/w) solution
of 0.1
and
and a 1.23: 1 (w/w) mixture (S) of hexadecyltrimethylammonium bromide
and
A pseudotemary phase map contained a significant microemulsion
A and B (60: 35: 5 and 20: 10: 70 : W: 0 , respectively) were used for reduction of several
region,
monocarbonyl compounds
(la),
ketone
acetophenone
and
an
and a
at 26”. For comparison purposes, reductions were also
performed in aqueous
A and
B) prepared by the substitution of
for the and 0 components of
A and B. Generally, the reductions were slightly faster in the
microemulsion media than in the corresponding aqueous
media. The significantly slower
reduction of relative to that of in
B indicated that the interphase is the reactive site. With
the influence of microemulsions on the competition between
and
was determined.
A and B there
In
A and B there was 8% and 11% 1,4-reduction, respectively, whereas in
was only a trace. With
the reactivity of the aromatic carbonyl group relative to that of the
aliphatic carbonyl group increased on going from
the sodium borohydride reduction of ketones, microemulsion catalysis is more effective than phase transfer
catalysis or the use of a tetraalkylammonium borohydride in a hydrocarbon solvent.
A and B to
A and B, respectively. For
Microemulsions are optically clear or translucent, basis. The unhatched region corresponds to micro-
thermodynamically-stable liquid dispersions of water
emulsions, and the hatched region to heterogeneous
that are mixtures. The microemulsions were stable for periods
in oil (w/o) and of oil in water
stabilized by a surfactant and, in most cases, a
cosurfacant such as a short-chain aliphatic alcohol.
of up to and beyond
h with respect to
hydride decomposition as determined by the lack of
They can
considerable amounts of a wide hydrogen evolution (visual inspection).’
Microemulsion compositions A and B
: 0, respectively) were
used for reductions at 26”. Conductivity sug-
gested that the former is an o/w and that the latter
is most likely a system. For comparison pur-
poses, reductions were also performed analogously in
aqueous solutions A and
B) prepared by the substitution of
variety of materials ranging from nonpolar organic to
ionic, inorganic compounds; moreover, they can si- (60 35 5 and 20 10 70
:
multaneously
compounds from both ends
hold poten-
for homogeneous reactions of organic
substrates with ionic inorganic reagents. Indeed,
of this spectrum. Thus,
tial as
there have
several reports of such reactions in
but very few of them have had
synthetic objectives. Holt and Gonzalez have studied
for the and 0 components of
A
and
reactions and we a
and B. Each reduction described below was followed
displacement reaction.’
with time by the direct analysis of the reaction
The
solvents for sodium borohydride reduc- mixture itself by reverse phase high performance
tions in terms of its solubihty and reactivity include
liquid chromatography (HPLC) with calibrated ul-
traviolet detection.
water, methanol, ethanol,
containing a
and
However, many rela-
An initial survey of borohydride reduction was
performed with monocarbonyl compounds
tively nonpolar organic compounds are insoluble in
these solvents. It is for such compounds that micro-
emulsions are attractive reaction media. Herein we
report the formulation of microemulsions containing
sodium borohydride, their use in the reduction of
several carbonyl compounds, and a determination of
the microemulsion reactive site.
phenone (la), benzaldehyde
acetophenone
eqn 1). The initial
to substrate was 2: 1 in A
and
molar ratio of
and
A and 1: 2 in
B and
B.
AND
Microemulsions were prepared from mixtures of
hexanes (0), a 50: 1 (w/w) solution
of 0.1 M
and a 1.23 1 (w/w) mixture (S) of
hexadecyltrimethylammonium bromide
and
la,
lb,
R = H
R = H
R =
The pseudotemary phase diagram
was constructed by titration techniques at 26.0 0.1”
and is given in Fig. 1; compositions are on a weight
R =
2691

2692
D. A.
et al.
0
W
Fig. 1.
pseudoternary phase diagram
= hexanes; = 50:
1
1
;
=
1.23:
1
(w/w)
The results for the reduction of la and 2a to molecular weight, slightly polar substrate that would
benzhydrol (lb) and benzyl alcohol
are listed in Table 1. For la, the data indicate that the media used for
reaction was slightly faster in A than in the case for in
in A. However, oil-rich
respectively,
be expected to have limited solubility in conventional
Indeed, this was
B and even
A and
A, and that it proceeded at about the same rate in
B solubilized and
B and
B. Direct comparison of the results allowed its reduction at the same concentration used
obtained in
B and
A and
A with those in
for la,
and
but at a rate about 18 times less
than that for 3a.9 It is proposed that this difference
results from the structural character of
B cannot be made because the molar
to substrate was different in each set
ratio of
(see above). For
the reductions were very fast in emulsions and reflects the nature of the reactive site.
Figure 2 depicts an idealized microemulsion; curva-
all four media with no detectable rate differences.
The results for the reduction of
(3b) and
and 4a to
ture of the assumed spherical dispersed droplets has
ignored. In general, water,
and
are
respectively, are given in Table 2. For
the reduc-
confined to the aqueous pseudophase, hexanes to the
to the interphase. The
and are distributed
between the aqueous pseudophase and the
tions were slightly faster in
A and B than in oil pseudophase, and
B, respectively. The results counterions OH-,
obtained with 4a are noteworthy. It represents a high
A and
Table 1 . Reduction of la and
in microemulsions and aqueous
8 8
9 1
9 9
9 0
9 3
6 0
9 0
1
1 0 0
1 0 0
1 0 0
1 0 0
2
3
1 0 0
of
an d
0 .0 9 1
in
A an d
A an d 0 .1 1
in
B
direct calibrat ed
an alysis; valu es are
du plicat e
an d
B.

2693
Reactions in microemulsion media
Table 2. Reduction of
and
in
and aqueous
2
-
m
UEB
71
94
52
95
15
20
59
25
85
89
96
55
83
115
290
1385
91
data for
in parentheses.
in B and
concentration of
0.092
in
A
and
A
and 0.11
B.
concentration of was 0.11
in
B.
direct calibrated
analysis; values are for duplicate
for and to for
and
are estimated to he
transesterification to n-butyl benzoate as determined
the oil pseudophase.” Since 3a and 4a are essentially by HPLC analysis. Even when potassium hydroxide
and
between the interphase and
phase,”
was deleted from
A, analogous results were
insoluble in water, there are two limiting possibilities
for the reactive site in
B, the oil pseudophase and obtained, so the reduction of 5 was not studied
further.
the interphase. Reduction in the former would be
expected to occur by a phase transfer process in-
volving HTABH,. The reaction of 3a with HTABH,
in benzene and
yield and 59% of 3b after 4 h and 2 h, respectively,
with the use of excess HTABH,. From
Table 2, it can be seen that after 25 min in B, 3a
at 25” has been reported” to
was 85% reduced. Thus it is very likely that not the
oil pseudophase but the interphase is the reactive site,
and the substantially greater reactivity of 3a over 4a
supports this conclusion. On solubilization in a
croemulsion, neutral organic compounds such as
and should partition between the oil pseudophase of
and the more polar interphase. Due to the greater in
The
ketone 6a and
7a
were used in two studies of
hydride reductions in
in
A and B. The molar ratio
to 6a in E A and
B and B was 1 2. Both
A was 2
and
and
and ketone
relative lipophilic character of
a lesser fraction of l&reduction can occur to yield
it should be solubilized at the interphase than of
respectively (eqn
Under the conditions used,
was monitored by the formation of
Thus, the much slower reduction of
is consistent
with the model of Fig. 2 and strongly suggests that
the interphase is the reactive site; on a time-averaged
basis, there is simply less 4a than 3a at the
alcohol
rapid reduction to
because
was undetectable due to its very
as demonstrated in a control
(see Experimental). The results are summarized in
Table 3. It is interesting to note that
Others have
similar reactivity
trends in
microemulsions for base-catalyzed hy-
occurred in
A and
A and B to greater extents than in
B, respectively. Nikles and
drolyses of pnitrophenyl alkanoates.
A preparative reduction of 4a in
B was per-
Sukenik have performed a similar
6a in aqueous micellar
with
formed under the conditions used for the kinetic runs
of Table 2. Analysis of the reaction mixture by HPLC
after four days indicated a 98% conversion of 4a to
and found 13% and
and micellar
21%
conditions,
under
4b. The isolation of
initial removal of
in 93% yield involved the
by the addition of KPF, to
l
.
OH-
1 - b u t a n o l
precipitate HTAPF,.
O H - +
The reduction of methyl benzoate (5) to benzyl
alcohol (2b) (eqn 2) was attempted in
A at 26”
to 5. However, not
.
1 - b u t a n o l
with a 2 1
ratio of
unexpectedly the reduction was accompanied by
saponification to potassium/sodium benzoate and
Pig. 2. Schematic representation of
idealized
sion.

2694
D. A.
in
et a l.
Table 3. Reduction of
and aqueous
90
52
33
20
10
48
67
76
24
24
71
44
92
92
t r
8
71
29
5
45
5
100
90
t r
12
3
85
89
180
1200
11
100
8
of
was 0.090 in A and
A and 0.11
analysis; values are for
are
in
B and
direct calibrated
and
B.
by
are estimated to
c = c
0
, c = c
1
6b
6d
As noted above, the results with 3a and 4a strongly
that of the aliphatic carbonyl group was enhanced by
a factor of about two in the microemulsion. The same
suggest that the interphase is the predominant site of
borohydride reduction in
B, and the same is conclusion can be drawn from the results in
assumed for
A. Therefore, in principle it should
B and
B. For example, in Table 4 consider the
be possible to reduce selectively one carbonyl group
product analysis at 2.5 min in
A. Of the aromatic
of a
if this group is preferentially solubilized carbonyl groups originally present in
32% of them
in the interphase on a time-averaged basis. To study
+ 28%
were reduced at this time, and
the possibility,
7a was used. The aromatic
carbonyl group of 7a may prefer the interphase as a
solubilization site more so than does the aliphatic
carbonyl group; it has been
other systems that aromatic groups absorb at
water interfaces.
with
Reduction of
and A with a molar ratio of
of 4
1: 1. Reduction products ketol
were obtained
7a was performed in
A
to 7a
and in
B and
B with a ratio of
and diol
and the results are given in
Tables 4 and 5. The regioselective reduction of the
aromatic carbonyl group was not realized. However,
a comparison of the reactivities of the two groups in
A with those in
A indicates that the
reactivity of the aromatic carbonyl group relative to

Reactions in microemulsion media
2695
Table 4. Reduction of
in microemulsion A
aqueous
solution A
70
3
12
2
2.5
S
28
4
63
53
42
10
4
4
2
15
56
22
12
23
44
15
60
90
90
96
of
0.045
solution.
calibrated
analysis; values are for duplicate
and are
to be
amount
2%) is in dicat ed bytr.
Table 5. Reduction of
in microemulsion
B
and aqueous
solution B
5640
3
3
1
2
a
44
65
2.5
15
31
1
5
.a
64
36
45
35
120
15
a
92
direct
and are
concentration of
0.033
in each solution.
calibrated
analysis; values are for duplicate
amount
3%.
the value for the aliphatic carbonyl groups was 91%
115 min, respectively, whereas under phase transfer
conditions with trioctylmethylammonium chloride in
(7h) + 28%
Thus, the relative reactivity
In A at 2 min the con-
ratio was 0.35
benzene/water at room temperature
was
This comparison is
versions were 15% and 82% for the aromatic and
aliphatic carbonyl groups, respectively, and the ratio
only reduced after 390
even more indicative of the greater reactivity of the
microemulsion systems when one considers the fact
was 0.18 1. At 5
the ratios were 0.47: 1 and
A, respectively. In
0.23: in
A and
that
is expected to be intrinsically less reactive than
in borohydride reduction due to the
B the ratio was 0.22:
1
at 2.5 min and in
B
was 0.10: at 2min. At 15 min, the ratios were
conjugating phenyl
0.44: 1 and 0.24: 1 in
spectively.
B and
B, re-
There have been other studies of carbonyl reduc-
tion with tetraalkylammonium borohydrides in
A comparison of the borohydride reductions above
with those performed under phase transfer conditions
with typical quatemary ammonium and
tic media other than that mentioned
et have used tetraethyl- and
Raber
monium borohydride in dichloromethane; under
these conditions diborane (produced by a reaction of
borohydride with dichloromethane) participates in
reduction. In benzene at room temperature, four
phonium
indicates that the former are much
faster. For example,
plete reduction in
underwent essentially com-
A and B in 20min and

D. A. JAEGER et
2696
that was recrystallized from hexane (
m.p. 65.8-66.5”).
to yield
m.p.
equivalents of tetrabutylammonium borohydride
effected less than 3% reduction of 3a during 36
Apparently, the practical use of such borohydrides in
hydrocarbon media requires the presence of a
solvent or other electrophilic
In conclusion, for the sodium borohydride reduc-
tion of ketones, microemulsion catalysis is more
effective than phase transfer catalysis or the use of a
tetraalkylammonium borohydride in a hydrocarbon
solvent. Furthermore, microemulsions are especially
attractive media for the reduction of relatively non-
polar substrates.
The reduction of
with
in ether at 10” by a reverse addition
yielded (83%) crude
as an oil, which by
NMR and IR
contained
material. This
oil was subjected to preparative HPLC
flow rate, 21
45 55 (v/v)
retention times:
8.5 min;
min]. The combined fractions containing
were rotary-evaporated to remove
aqueous residue was extracted twice with ether. The com-
bined extracts were dried over and
evaporated to give 3638” (lit
which was homogeneous by the HPLC anal-
ysis conditions used for reactions of (se-e below).
and the
EXPERIMENTAL
procedures. All melting and boiling points are
The
give (38%) 7a:
of
Lansbury and:
3637” (lit”
to
was
uncorrected. The
NMR spectra were recorded with a
NMR 6 7.70-7.82
4 H.
JEOL FX-270 spectrometer
and
was
H,
7.24-7.33 (m,
used as solvent with TMS as internal standard. Infrared
spectra were obtained on a Beckman Model IR-IO
trophotometer with neat or Nujol mull samples between
plates. Mass spectra were recorded on a Kratos
MS-50 spectrometer at the Midwest Center for Mass
trometry at the University of Nebraska-Lincoln. The ioniz-
2 H,
2.98
(s).
3 H,
IR
(s) anh
The procedure of
Lansbury and Peterson” was used. Reduction of
7a gave (82%) an oil which by NMR analysis was pure
However, by the HPLC analysis conditions used for
ing voltage was 70
the filament current 500
and the
ion-source temperature 250”; direct insertion was used with
ambient probe temperature. High performance liquid
chromatography analyses were performed on two
stainless steel columns packed with
reactions of
detected. The oil was subjected to preparative HPLC
ant, 55 45 (v/v) flow rate, 16.8 reten-
(see below), an unknown impurity was
tion times:
9.0 min; impurity, 6.0 min] to give as an
column A, 4.6 mm (id) (Altex), and column B, 4.0 mm
(id) (EM). Beckman Model 332 and 344
oil:
NMR 7.237.84
9 H,
3.86 fsextet,
IR
J = 7 Hz, H, OCH),
(m, H,
were fitted with a column inlet filter (2
cm x 4.6 mm
between the
and a
3 H.
and
1.25
J = 7 Hz.
2960
1408
and column. For detection and (neat)
3050 (w),
1370
quantitation, Beckman Model 153 (254nm) and 165
(variable-wavelength) ultraviolet-visible detectors and a
Hewlett-Packard Model 3390A reporting integrator were
1650 (s), 1600 (s), 1570 (w), 1440
1060 (wj,
780
725
1010
928
(s).
913
830
mass
used. A backpressure regulator was attached to the outflow 690
line of each detector. Preparative liquid
for
254.1307.
separations were performed on a 25 cm x 21.2 mm (id)
stainless steel column packed with
Pont). The Beckman Model
Zorbax ODS (du The procedure of Lansbury and Peterson” was used. Re-
equipped
duction of
contained
gave an oil which by ‘HNMR
in addition to and
with preparative pump heads, was used with a column inlet
filter
between the sample injector and column. The
equipped with a preparative
diol 7d. This material was subjected to preparative HPLC
Beckman Model 153
[eluant, 55: 45 (v/v)
retention times:
5.5 min] to give
flow rate, 16.8
12.5 min; 9.0 min;
as an oil: NMR 7.1 l-7.42
Hz, H,
H, OCH), 2.86
flow cell, was used for detection. For all analytical and
preparative HPLC chromatography, mixtures
7.0 min;
and
and
Sodium borohydride (Alfa,
T.
were used at the
J
=
7
2
9
H,
(s,
1
indicated below.
powder) was
I R
used as received, and HTABr (Aldrich) was recrystallized
1490(w),
1360
254.1303;
twice from 4: (v/v)
(25”). The
(Mallinckrodt,
1175 (m), 1020 (m), 790 (w), 720 (w), 690 cm-’ (s). High
resolution mass spectrum:
254.1307.
anes (J. T. Baker, HPLC grade),
spectral grade), and
for
T. Baker, HPLC grade)
were used as received. Benzophenone(la) was recrystallized
I n
in
from hexane (25”) to give material with m.p.
and
standard fashion.
ether to give
was reduced with
benzhydrol (lb) from 4: (v/v) ether-hexane (25”) to give
which contained a trace of
material with m.p. 68-69”. Benzaldehyde
h o l
alco-
a n d
by IR analysis. This material was purified by preparative
HPLC [eluant,
(v/v)
flow rate,
9.0 min] to give
methyl benzoate (5) were purified by distillation. The fol-
lowing materials (Aldrich) were used as received:
16.8 retention times: 5.7 min;
diol 7d as an oil (mixture of racemates):
phenyl-3-buten-2-one
none
m.p.
7.08-7.48 (m,
5.78 (s, H,
3.77 (sextet,
overlapping with m,
overlapping with
J = 7 Hz, 3 H.
and
used
B
Y
the HPLC
(see be-
J = 7 Hz, 1 H, OCH), 2.31-2.79
of
3 H total, OH and
1.42-1.80
low), all materials were homogeneous except for and
m, 3 H total, OH and
1.18
which contained a trace of
and an unknown impurity,
(w), 2920
IR (neat) 3350
3050
1485
respectively, that were undetectable by NMR.
1370
1170
1600
1505
1445
The preparation of
followed an established
Crude material was
Pseudoternary phase diagram. A pseudotemary phase
diagram (Fig. 1) of surfactant, aqueous, and oil components
was constructed at 26.0 0.1” to give a region of clear,
homogeneous solutions. The surfactant component (S) con-
recrystallized once from 1: (v/v)
twice from
(
and three times from ether
58-59” (lit m.p.
In standard fashion, re-
in ether gave a white solid
to give
sisted of a
(w/w) mixture of HTABr and
duction of
with
respectively, and the oil component (0) of hexanes. The

Reactions in microemulsion media
2697
spectively;
spectively.
24: 1, respectively;
respectively; and
aqueous component (W) consisted of 50: 1 (w/w) 0.1 M
With a variation of S/W from 85 15 (w/w)
re-
to 10: 90 by increments of 5, samples of and W were
weighed
6 g total) into ground-glass stoppered bottles.
Control on HPLC analyses. To a 5.0 g aliquot of
A
of
a
Each sample mixture was then titrated with hexanes to give
the points at which it turned from heterogeneous (cloudy)
at 26” was added
7a to give a 0.046
After 2.5
to homogeneous (clear) and then back to heterogeneous sample was withdrawn and analyzed by HPLC with the
(visual observation). The resulting points were used to same conditions used for the other reductions of above
(Table
except that the flow rate was
Other
detemine the rough boundary of the microemulsion region,
and thirty samples (cu. 1 g each) around this boundary were
equilibrated at 26” for 48 to 72 h. They were then titrated
identical reaction mixtures were prepared and analyzed
similarly at 2.5 min but with flow rates of 1 and 2
The results of the aoalyses at the three different
were within
‘rates
with either 0 or
W with equilibration at 26” for 48 to 72 h
of one another. Thus, it is apparent that
between additions. The resulting points gave Fig. 1.
Conductivity measurements. The following conductivities
were measured with a Yellow Springs
minimal reduction occurred within a sample of a reaction
mixture subsequent to its injection into the HPLC
matograph.
ment
Model 31 conductivity bridge equipped with a
Preparative reduction of
To
YSI Model 3403 conductance cell with a cell constant of
2.5 g of
B at
was added 93.5 mg (0.270 mmol) of
1
A, 11,800;
B, 57; hexanes
grade),
(dis-
to give a concentration of 0.11 m with
to molar
0.2;
tilled from
1 (w/w) 0.1 M
ratio of 1: 2. The reaction mixture was held at 26” for 4 davs
and was then analyzed by HPLC with the same conditions
then boiled). 0.2.
Reductions. The following general procedure was used. To
used for the reduction of
of to was complete. To the reaction mixture,
which contained 0.279 g (0.766 mmol) of
ded a solution of of KPF, in
A white precipitate of
and the mixture was shaken well and allowed to stand for
30min. It was then filtered twice through a fritted-glass
funnel, and the solid in the funnel was washed with two
ml portions of hexane. The aqueous layer was extracted
three 25 ml portions of hexane. All hexane layers were
above (Table 2). The reduction
a
50ml glass-stoppered round-bottomed flask was added
2.5 5.0 or 10.0 of the reaction medium
A and
was ad-
of
B a&l
brated at
A
B), which was-then equili-
(at least 90 min for
B and at least 45 min
formed immediately,
for the others). Then the appropriate amount of substrate
was added to give: a 2 1 molar ratio of
to substrate
for la,
and in
A and
A; a
B and
molar
ratio for the same substrates and
in
a 4: 1 molar ratio for
in
A and
A; and
B. Samples
a 1: 1 molar ratio for in
B and
combined and dried over
left 87.6 mg (93%) of
neous by NMR analysis.
and rotary-evaporation
which was homoge-
(5 to
of the reaction mixture were then removed by
those
syringe at appropriate times and analyzed by
aliquots that could not be analyzed immediately were added
to melting point capillaries, frozen at once in liquid N,, and
then stored at -10” until analyses could be made. For
of I-phenyl-1-octaakcanone (4a) in various
media. The approximate solubilities at 26” of in 2.5 g
aliquots each of
prepared without
A,
A, and
B, all
“complete” reaction corresponded to
95% of substrate. All reaction
comparison purposes,
were determined by visual in-
the disappearance of
spection. Each medium was equilibrated at 26” for at least
mixtures remained homogeneous throughout the time
1 h before the addition of
shaken and allowed to stand at 26” with intermittent
shaking. For A, an amount of equivalent to of
the molar amount of used in this medium (Table 2) was
added, most of it remained undissolved after 24 h. When a
equivalent of was added to A, most of it
and then the mixture was
needed for complete reaction except that for
in
this system was slightly turbid immediately after
the addition of 7a and became Very turbid (without phase
separation) after 30 min.
The HPLC analysis conditions were as follows: (sub-
strate: HPLC column, detection wavelength,
com-
dissolved after 5 h, but it was still incompletely dissolved
position, flow rate, retention times of substrate and pro-
ducts): la: column A, 254 nm, 65 35 (v/v)
after 24 h. For
of the molar
A, an amount of
to
3a used in this medium was
1
la, 8.0 min, lb 6.0 min;
column A, 254
8.5 min,
added; most of it remained undissolved after 24 h. For
45: 55 (v/v)
1
B, an amount of
amount of used in this medium was added; most of it
remained undissolved after 24 h. Wheo a molar equiv-
alent of was added to B, some of it dissolved
after 5 h, but it was still incompletely dissolved after 24 h.
equivalent to
of the molar
6.0 min;
1 ml/min,
column A, 254 nm, 45 55 (v/v)
8.5 min, 6.5 min;
25.5 min,
column B, 254 nm,
2 ml/min,
24.0 mio;
column B,
33.0 min,
220 nm, 20 80 (v/v)
35.5 mio,
2
7a: column B,
7a. 11.6 min.
The data for the various
254
9.0 min,
reductions are listed in Tables l-5.
The reduction of methyl benzoate (5) was attempted using
the above procedure. To 10 g of
5 to give a 0.091 m solution. Uncalibrated analysis by reduction to give alcohol
55 45
1
Control on the reduction of
A 0.10 m solution of
in
B at 26” was prepared (molar ratio of to
7.
= 0.5 1). At 5 mio and 1410
analysis by HPLC with
the conditions used for the reduction of
(see above)
A was added 0.124 g of indicated the presence of only Thus, does not undergo
Control on the
reduction of 7.9
performed at
above. The initial concentration of
approximately to that of alcohol present after complete
of
The
B was
procedure
HPLC [column A, 254 nm 65: 35 (v/v)
at various times indicated the presence of
tial amounts of sodium (potassium) benzoate (retention
time, 3.8 mio) and n-butvl benzoate (11.4 in addition
to that of An analogous run
1
(0.053
of in 4.85 of
the general
0.011 m, corresponds
5
was performed in a microemulsion prepared without KOH.
Lesser amounts of sodium benzoate and n-butyl benzoate
were detected.
reaction of
molar ratio of
was reduced to
in
B (Table 3). Therefore, the initial
to was 5: 1. After 10 min, 89% of
and a second run under identical
Calibration factors for HPLC analyses. For the HPLC
conditions employed for the analyses of reaction mixtures
(see above), detector response factors relating each substrate
to its reduction product(s) were determined with the use of
known mixtures of these materials: la-lb, 51: 1, re-
conditions gave the same result. Thus, is much more
reactive
and it is reasonable that it does not
accumulate to allow its detection by HPLC.
made to the U.S. Army Research Office
and to the Marathon Oil Company for support of this
spectively;
123 1, respectively;
64 1,

D. A. JAEGER et al.
research, and to the National Science Foundation for a
additional modest decrease in intrinsic reactivity is ex-
pected. In a study of the hydroxide ion-catalyzed hydro-
lysis of a series of p-nitrophenyl esters in water at
departmental grant (CHE-8026553) for the purchase of a
JEOL FX-270 NMR spectrometer.
Guthrie [J. Guthrie, J. Chem.
Chem. Commun. 897
found that the decanoate ester was only 4.3 times
less reactive than the acetate ester. The rate decrease was
ascribed to coiling of the alkyl group back on itself with
a resultant shielding of the ester carbonyl group. For
in the oil pseudophase, the coiling
because an alkyl group is reasonably expected to be more
extended in this medium than in water. For at the
interphase, the coiling effect likewise is less important
RE FE RE NCE S AND NOTE S
‘For the previous paper in this series, see C. A. Martin, P.
M.
G. H. Angelos and D. A. Jaeger, Tet-
is less important
rahedron Letters 23, 4651 (1982).
M. Prince,Microemulsions: Theory and Practice. Aca-
demic Press, New York, (1977); Danielsson and B.
Lindman,
3, 391 (1981).
because. of its probable orientation with the
group
‘For examples, see R. A.
Ado.
Interface
in the interphase and the alkyl group extended into the oil
pseudophase.
Sci. 15,
and Refs. therein.
Gonzalez and S. L. Holt.
Chem. 46.2594 (1981).
Mandohni,
reduction did occur by a phase transfer process in the
For a comment, see C.
oil pseudophase, one might expect
to be more reactive
Chem. Commun. 251 (1982).
than 3a due to its greater concentration therein.
Gonzalez and S. L.
Org. Chem. 47, 3186 (1982).
M. Menger, J. A. Donohue and F. Williams, Am.
0. House, Modern Synthetic Reactions 2nd Edn, p. 47.
Benjamin, Menlo Park, California (1972).
Chem.
286 (1973);
F.
S. L.
526 (1979).
and C.
have re-
and
R. E.
J.
and
Interface
‘Hydrogen evolution was detected immediately from
“van
G. van
1301
50: 1 (w/w)
0.01 M
1437
with
croemulsions
Koningsberger,
ported catalysis of
of the hydroxide ion-catalyzed
D. Dreher and R. D. Sydansk, Pet.
hydrolysis of the related
ketone
(1971).
An analogous hydrolysis of
rate comparison was based on the average one-point
second order-rate constants for and calculated
C. Brown. 0. H. Wheeler and K. Ichikawa. Tetrahedron
6a in the present study did not occur as demonstrated by
the absence of
alcohol.
and its reduction product,
1,214
from the data of Table 2 at times 25 min and
A Nikles and C. N.
4211’ (1982).
Tetrahedron Letters
55 min. For each calculation, the initial concentration of
ketone was assumed to be 0.1 M and that of borohydride
0.05 M.
a comprehensive study of
vs
of
with sodium borohydride, see M. R. Johnson and
‘@The binding of counterions to the interphase of a
emulsion may be similar to that of counterions to the Stem
B. Rickbom, J. Org. Chem.
1041 (1970).
H. Fendler and E. J.
and Macromolecular Systems, Chap.
examples, see
Catalysis in
Layer of a
S. Romsted,
(Edited by K. L.
bilization, and
3. Academic Press, New York (1975);
H. Fendler
Vol. II, p. 509. Plenum Press, New York
“The oversimplified model of Fig. 2 neglects the presence of
small amounts of various components in regions other
than those indicated (i.e. water of hydration for ionic
species in the interphase).
Membrane Mimetic Chemistry, Chap. 2.
Interscience, New York, and Refs. therein.
*‘This and subsequent ratios reflect the reactivity of the
aromatic carbonyl group in both 7a and
and that of the
at time
aliphatic carbonyl group in 7a and
A. Sullivan and A. A. Hinckley,J. Org. Chem. 27, 3731
(1960).
will the ratio represent the reactivities of the carbonyl
groups in alone.
different lipophilic characters of the n-alkyl groups of
M. Starks and C. Liotta, Phase Transfer
PP.
and
are considered relative to the slightly polar
317-324. Academic Press, New York
character of the common benzoyl group.
Colonna and R. Fomasier.
J. Raber and W. C.
531
Org.
Solution Chemistrv of Surfactants (Edited bv K.
690
L.
Vol. I, p. 153.
(1976);
J. Raber, W. C. Guida and D. C.
has determined that within an aqueous
even a very
berger, Tetrahedron Letters
5107 (1981).
slightly polar (surface active) substance such as benzene is
preferentially at an interface rather than in the
hydrocarbon core. It is assumed that analogous effects
operate in microemulsions.
6, p. 51.
A. Mikeska, C. F. Smith and E. Lieber, J. Org. Chem.
2, 449 (1937).
Claus and H. Hafelin. J.
Chem. 121. 54. 399
“It is realized that there is a difference between the intrinsic
reactivities of and with respect to reduction by
(1896).
For example. Brown and Ichikawa
C. Brown and K.
found that “H.
with (1957);
at 25”. This difference is due
to steric effects since. all alkyl groups have about the same
polar effect Char-ton, J. Am. Chem. 97, 3691
As the alkyl chain length increases to 17 in an
L. Breusch and M.
Ber. 87, 1225 (1954).
Am. Chem.
and P. Perrin.
Chim. Fr. 801
phenyl ethyl ketone is 1.8 times less reactive than
Grummitt, Organic Syntheses, p. 771, Coll.
respect to
in
Vol. IV. Wiley, New York
A. Hochstein and W. G. Brown, Am. Chem.
3484 (1948).
70,
T. Lansbury and J. 0. Peterson, Ibid.
2236 (1963).